Update, 9 June 2021: Google reports this week in the journal Nature [nature.com] [DOI: 10.1038/s41586-021-03544-w] [DX [doi.org]] that its next generation AI chip, succeeding the TPU version 4 [ieee.org], was designed in part using an AI that researchers described to IEEE Spectrum last year. They've made some improvements since Spectrum last spoke to them. The AI now needs fewer than six hours to generate chip floorplans that match or beat human-produced designs at power consumption, performance, and area. Expert humans typically need months of iteration to do this task.
Original blog post from 23 March 2020 follows:
There's been a lot of intense and well-funded [ieee.org] work developing chips that are specially designed to perform AI algorithms [ieee.org] faster and more efficiently. The trouble is that it takes years to design a chip, and the universe of machine learning algorithms moves a lot faster than that. Ideally you want a chip that's optimized to do today's AI, not the AI of two to five years ago. Google's solution: have an AI design the AI chip.
"We believe that it is AI itself that will provide the means to shorten the chip design cycle, creating a symbiotic relationship between hardware and AI, with each fueling advances in the other," they write in a paper describing the work that posted today to Arxiv [arxiv.org].
"We have already seen that there are algorithms or neural network architectures that... don't perform as well on existing generations of accelerators, because the accelerators were designed like two years ago, and back then these neural nets didn't exist," says Azalia Mirhoseini, a senior research scientist at Google. "If we reduce the design cycle, we can bridge the gap."
Related: Google Reveals Homegrown "TPU" For Machine Learning [soylentnews.org]
Google Pulls Back the Covers on Its First Machine Learning Chip [soylentnews.org]
Hundred Petaflop Machine Learning Supercomputers Now Available on Google Cloud [soylentnews.org]
Google Replaced Millions of Intel Xeons with its Own "Argos" Video Transcoding Units [soylentnews.org]